Whatfix

What it is

Whatfix is a digital adoption platform: a JavaScript layer installed over software you did not build — Salesforce, Workday, SAP, an EHR, a homegrown claims system — that renders walkthroughs, tooltips, task lists and a self-help panel inside the host application. The category reference is WalkMe, which SAP bought for about $1.5B and closed on 12 September 2024. That deal is why this page exists: Whatfix is now the largest DAP not owned by an application vendor with a stake in which applications you adopt.

Five products sit on one contract: DAP for web and desktop, Mobile, an operating-system layer for Windows and Mac, Mirror (screen-capture training simulations with AI roleplay), and Product Analytics. Whatfix raised a $125M Series E led by Warburg Pincus with SoftBank Vision Fund 2 in September 2024, at a reported $900M valuation and roughly $75M ARR, and claims 700+ enterprise customers including 80+ of the Fortune 500.

Why it shows up in Customer Success stacks

• Guidance rides on top of applications you do not control. Pendo, Userpilot and Appcues instrument a product your own engineers ship. Whatfix takes the opposite job: your CSMs, or your customers’ employees, are stuck in someone else’s UI and no code change is coming. When the adoption problem lives in Workday or in a customer’s ERP instance, an overlay is the only lever you hold.
• ScreenSense is the authoring economics. Introduced in 2025, it is a computer-use model trained to read GUIs, and three agents shipped on it on 9 September 2025: Authoring (build a walkthrough from a prompt), Insights (ask analytics questions in prose), Guidance (answer questions inside the workflow). Flow authoring is the recurring cost in every DAP, so the agent that writes flows matters more than the one that answers questions.
• Mirror covers training you cannot sandbox. Recorded simulations of the real screens are what you hand 400 people when giving them a live environment is not an option — the gap between a training deck and something that transfers.

Pricing reality

Whatfix publishes no dollar figures. The meter is a flat platform fee plus per-license fees: employee-facing applications priced on total headcount with access, customer-facing applications on monthly active users, and every application you cover is its own line.

Third-party contract data puts standard agreements in the $23,700–$37,100 per year band, averaging near $32,000 — the realistic entry point for one application. Multi-app Enterprise deployments run past $100K/yr. Implementation and content labor add a further $10K–$100K in year one (third-party estimate; Whatfix publishes no services rates), and that is the line most evaluations leave out of the business case.

What the tiers gate: Standard caps Self Help at 2,000 articles, ships out-of-the-box surveys only, manual translation, and two integrations. Premium removes the article cap and adds custom surveys, auto-translation, unlimited integrations, and automated flow testing. Product Analytics is included free alongside a DAP web plan but capped at 25 tagged actions; unlimited tagging is a Premium line. Auto-translation and automated flow testing decide whether a multi-locale rollout stays maintainable, so most real deployments are Premium, not Standard — price accordingly.

Best for

Enterprise IT, L&D, and Customer Success leaders rolling a purchased application — Salesforce, Workday, SAP, ServiceNow, an EHR — out to 1,000+ employees, or training a customer base you do not employ. ROI is strongest where the alternative has a price tag you can name: support tickets and instructor-led hours. At roughly $32K/yr for one application, a 5% deflection against a large population clears the number; against 200 seats it does not.

Do not buy it if

You are a product-led SaaS company onboarding self-serve signups into your own product. The ~$24K/yr floor and a 1–3 month implementation are an order of magnitude wrong against Userpilot, Appcues or Chameleon, which install in a day and start in the hundreds per month. Skip it too if SAP is the application you are adopting and you already hold SAP BTP entitlements — WalkMe now ships inside that platform, and paying separately for the independent option only pays off when your non-SAP footprint is the larger problem.

Versus the alternatives

• WalkMe — largest mind share in the category at 25.2% (PeerSpot, June 2026) and deeper SAP alignment by construction. Pick it when SAP is the system being adopted and the bundle economics are real. Pick Whatfix when the application portfolio is mixed and you do not want the adoption layer owned by one of your application vendors.
• Pendo — pick it when analytics is the primary job, guidance is secondary, and the application is one your engineers ship. Pendo’s guides sit beside product analytics you already need; Whatfix’s analytics exist to explain its guidance.
• Guidde — the fastest-growing entrant, on a $50M Series B led by PSG Equity in February 2026. Pick it when the deliverable is video documentation produced quickly by non-specialists rather than a governed guidance layer with segmentation, SSO, and per-role targeting.
• Userlane — pick it when you want walkthroughs live in days across a narrow application set and can accept a smaller footprint and community.

Watch-outs

• Flows break when the host application’s UI changes. Reviewers on G2 (4.6/5 across roughly 400 reviews) name this repeatedly, and it is inherent to overlaying software you do not control. Guard: fund one named content owner per major application from day one, buy Premium for automated flow testing, treat every host-app release as a regression event, and anchor step selectors to stable element IDs rather than DOM position wherever the host exposes them.
• Implementation runs 1–3 months, not weeks. Guard: scope phase one to a single application and under 10 flows, and capture the baseline you will be judged on — ticket volume for that application, task-completion rate — before go-live. Without the pre-measurement the renewal conversation has no evidence.
• Per-application metering makes the second application cost like the first. Guard: negotiate the multi-app Enterprise rate at initial signature, not at expansion. Once one application is live and its content is built, the switching cost is yours and your negotiating position is gone.
• No MCP server. Whatfix ships REST APIs (developer.whatfix.com, in beta), a Report Analytics API that returns CSV, and xAPI content export into an LMS on Enterprise accounts — but nothing an agent connects to natively, unlike Amplitude or PostHog in the analytics layer. Guard: if agent-driven reporting is in the plan, budget a wrapper over the Report Analytics API instead of assuming a connector will appear.
• Standard’s two-integration limit and manual translation are quiet blockers. Guard: if the rollout is multi-locale, or needs both a knowledge base and a CRM attribute source for segmentation, price Premium into the evaluation from the start. Finding this mid-implementation turns into a change order at the worst possible moment.